Analysis on Incorporating the Influenced Factors for Online Product Sale by Review System and Promotional Strategy using Optimization of Online Promotion System

Abstract

The goal of this study is to look into the role of online promotional marketing and online reviews in predicting consumer product demand. It is tried to forecast if online review factors like valence and volume of reviews, the frequency of positive and negative reviews, and online promotional marketing variables like discount, may affect product demand on Amazon.com using data from Amazon.com. Asynchronous Input / Output calls were used to scrape the Amazon.com sites using Node.JS agents. After that, the data sets from Web crawling and scraping were preprocessed for Deep Neural Network analysis.For the promotional strategy, this study used 12 characteristics, which were optimized using two optimization methods, PSO and Eagle. For both optimization techniques, the deep learning methods like CNN and LSTM were applied and the combination is PSO_CNN, PSO_LSTM, Eagle_CNN, Eagle_LSTM. This article suggested a hybridization of two Deep learning methods for both optimization techniques called WPCNLSTM (Weighted PSO CNN and LSTM) and WECNLSTM (Weighted Eagle CNN and LSTM). This study proved that the proposed algorithm WECNLSTMworked efficiently in finding the best feature for product promotional strategy with high accuracy of 80%, 79% sensitivity, 94% specificity, 82% F-Score,94% Precision and 78% Recall.